Machine Learning Interatomic Potential for Simulations of Carbon at Extreme Conditions

TL;DR

This paper introduces a machine learning interatomic potential for carbon that accurately simulates extreme conditions, enabling large-scale molecular dynamics with near-quantum accuracy.

Contribution

The authors developed a novel SNAP machine learning interatomic potential for carbon, validated against quantum molecular dynamics and experimental data, with high accuracy and transferability.

Findings

Predicts carbon phase diagram with 3% accuracy

Accurately models melting curves and shock Hugoniot

Enables large-scale atomic simulations at extreme conditions

Abstract

A Spectral Neighbor Analysis (SNAP) machine learning interatomic potential (MLIP) has been developed for simulations of carbon at extreme pressures (up to 5 TPa) and temperatures (up to 20,000 K). This was achieved using a large database of experimentally relevant quantum molecular dynamics (QMD) data, training the SNAP potential using a robust machine learning methodology, and performing extensive validation against QMD and experimental data. The resultant carbon MLIP demonstrates unprecedented accuracy and transferability in predicting the carbon phase diagram, melting curves of crystalline phases, and the shock Hugoniot, all within 3% of QMD. By achieving quantum accuracy and efficient implementation on leadership class high performance computing systems, SNAP advances frontiers of classical MD simulations by enabling atomic-scale insights at experimental time and length scales.